1 Recommendation

Client: NSW Police
This report analysed the trend of drunk driving through the data of Random Breath Testing (RBT), and reviewed the effectiveness of RBT on deterring drivers. The result suggests that the RBT is effective and can be further implemented.

2 Initial Data Analysis

The data of Random Breath Testing and road casualties related to drunk driving in NSW is extracted from The Australian Random Breath Testing on the BITRE. The data recorded in NSW is extracted and the percentage of positive RBT over the total number of RBT conducted is calculated.

# Calculate new data
RBT3_NSW =  mutate(RBT2_NSW, percent.positive = (Positive.RBT/RBT.conducted)*100)
RBT3_NSW$percent_positive = as.numeric(RBT3_NSW$percent.positive)
str(RBT3_NSW)
## 'data.frame':    12 obs. of  9 variables:
##  $ Year            : chr  "2008" "2009" "2010" "2011" ...
##  $ State           : chr  "NSW" "NSW" "NSW" "NSW" ...
##  $ RBT.conducted   : num  4204525 4440862 4637033 4520010 4735462 ...
##  $ Positive.RBT    : num  27368 26595 24411 22117 19982 ...
##  $ Licences        : chr  "" "" "4,791,490" "4,893,688" ...
##  $ driver.killed   : int  58 68 48 52 44 46 34 36 41 42 ...
##  $ total.deaths    : int  78 94 74 70 56 53 50 45 59 55 ...
##  $ percent.positive: num  0.651 0.599 0.526 0.489 0.422 ...
##  $ percent_positive: num  0.651 0.599 0.526 0.489 0.422 ...
##  - attr(*, "na.action")= 'omit' Named int [1:36] 13 14 15 16 17 18 19 20 21 22 ...
##   ..- attr(*, "names")= chr [1:36] "13" "14" "15" "16" ...

3 Evidence

3.1 Trend of drunk driving in NSW

This part focuses on the trend of positive RBT recorded in NSW from 2008 to 2019. The percentage of positive RBT recorded of each year is presented in the line graph below.

RBT3_NSW %>%
ggplot(aes(x = Year, y = percent.positive, group = 1)) +
  geom_line(color = 'black', size=0.8, alpha=0.9) +
  geom_point(size=2) +
  labs(title = "The percentage of positive Random Breath Testing (RBT) \n recorded in NSW by year", x = "Year", y = "Percentage of Positive RBT (%)") +
  theme(plot.title = element_text(hjust = 0.5, size = 16), axis.title.x = element_text(size=13), axis.title.y =
           element_text(size=14))

The line graph shows the decreasing trend of positive RBT recorded, although there is a slightly rebound from 2015. It is believed that drunk driving is generally decreasing in NSW.

3.2 The effectiveness of RBT

The scatter graph below jointed the data of RBT conducted each year and the number of drivers killed with a blood alcohol concentration (BAC) above the legal limit. The correlation coefficient between two data is then calculated to show the relationship between RBT and death count of drunk drivers.

RBT3_NSW %>%
  ggplot(aes(x = driver.killed, y = RBT.conducted)) +
  geom_point() +
  geom_smooth(method = "lm") +
  labs(title = "Relationship between total times RBT conducted and driver \n being killed with a BAC above the legal limit in NSW", x = "Number of driver died with an excess BAC", y = "RBT conducted (times)") +
   theme(plot.title = element_text(hjust = 0.5, size = 15), axis.title.x = element_text(size=13), axis.title.y =
           element_text(size=13))

The correlation coefficient calculated is -0.7, a negative relationship of numbers of RBT conducted and driver death, which means the more RBT conducted help reduce drunk driving incidents. RBT plays a role as a deterrent effect to the drivers that reminding them not to drink before driving. Therefore, this report reflects that the RBT has a positive effect on deterring drivers from drunk driving and it is supported to continue implementing.

4 Acknowledgments

Data of RBT retreived from https://data.gov.au/data/dataset/a814a8c5-ef57-463c-9c8b-a6e625bfb860/resource/6c5cbea3-79dc-40b9-9775-49521a57eacb/download/bitre_enforcement_data-rbt.csv

References
Terer, & Brown, R. (2014). Effective drink driving prevention and enforcement strategies : approaches to improving practice. Trends and Issues in Crime and Criminal Justice, 472, 1–8. https://doi.org/10.3316/agispt.2014270